Clinical Analysis And Prediction Model Of 546 Cases Of Solitary Pulmonary Nodules

Objective:To investigate the value of CT scan and tumor markers of solitary pulmonary nodule(SPN)in identifying SPN as benign and malignant,and to establish a prediction model for solitary pulmonary nodule in order to reduce human disturbance and provide suggestions for the treatment of SPN.Methods:Collect medical records of patients with solitary pulmonary nodule who have been treated by thoracic surgery and confirmed by pathology at the First Hospital,affiliated to the China Medical University.Collect their gender,age,smoking history,clinical symptoms(cough or expectoration,chest congestion or shortness of breath,chest pain or hemoptysis,hypothermia or weakness),nodular diameter,CT scan(nodular marginal state,pleura depression and vascular cluster syndrome)and tumor markers determination(CEA,NSE,CYFRA21-1,).The data collected were divided into malignant and benign groups according to pathological results,Comparison and analysis on gender,age,smoking history,clinical symptoms,nodule diameter and CT signs using SPSS 23.0software.At the same time,the independent risk factors of the malignant probability of solitary pulmonary nodule were selected by logistic regression analysis and the regression equation model was constructed.Result:In this study,546 cases of disease,393 cases of malignancy and 153 cases of benign disease were collected.There were significant differences between malignant group and benign group in age,sex,smoking history,nodular diameter,CT scan(nodular marginalstate,pleuradepressionandvascularclustersyndrome)and CEA,CYFRA21-1(P<0.05).And there was no significant difference between malignant group and benign group in linical simptoms,NSE.Multifactorial logistic regression analysis showed that age,smoking history,diameter,spurs,pleura depression,and vascular cluster were independent risk factors for SPN.The mathematical equations of benign and malignant predictions are:Y=e~x(1+e~x),X=-4.553+(0.045×age)+(1.076×smoking history)+(0.761×diameter)+(0.72×spurs)+(0.97×pleura depression)+(1.575×vascular cluster).e was natural logarithm.At cut-off point=0.5380,the approximate index was the largest(approximate index=sensitivity+specificity-1),where model sensitivity was 91.8%,specificity 73.0%,area under the curve 0.886,95%CI:0.823-0.948.Conclusion:It is difficult to judge the nature of SPN,and special attention should be paid to patients with age,smoking history,larger diameter of nodules,spurs,pleural depression,and vascular cluster syndrome.The mathematical prediction model of benign and malignant judgment of SPN established by logistic regression is more effective.